主分类：搜索推荐与业务决策算法研发
辅助：大模型与生成式算法研发（仅展示）
来源 JD 数据（旧分类保留，当前主分类见上方）
{
"schema_version": "via_ai_product_manager_jobs_v1",
"data_type": "ai_product_manager_jobs",
"job": {
"job_title": "基于预训练大模型的端到端生成式搜索技术研究-搜索（新加坡）Research on End-to-End Generative Search Technology Based on Pre-Trained Foundation Models-Search (Singapore)",
"job_level": "校招",AI生成
"work_category": "全职",AI生成
"department": "未披露",
"location": "新加坡",
"work_mode": "未披露",AI生成
"job_category": "算法"AI生成
},
"company": {
"company_name": "字节跳动",
"company_industry": "互联网 / 数字内容 / 电商",
"company_size": "10000人以上",
"company_website": "bytedance.com",
"company_description": "字节跳动提供内容创作、信息分发、电商和协作产品，旗下产品包括抖音、今日头条、TikTok和飞书。"
},
"compensation": {
"salary_range": {
"min": "",
"max": "",
"currency": "未披露",
"cycle": "未披露"
},
"benefits": "未披露"
},
"requirements": {
"working_years": "",
"degree": "获得博士学位",
"language": "未披露",AI生成
"visa_support": "未披露",
"responsibilities": "团队介绍：字节跳动搜索团队主要负责抖音、国际化短视频、今日头条、红果短剧、番茄小说、AI搜索等产品以及电商、生活服务等业务的搜索算法创新和架构研发工作。我们使用前沿的机器学习\\大模型技术进行端到端建模并不断创新突破，同时专注于分布式系统、机器学习系统的构建和性能优化，从内存、Disk等优化到索引压缩、召回、排序等算法的探索，致力于支撑字节跳动持续多元化的产品创新和高速的业务增长、基于大模型革新和重塑下一代搜索技术体系，充分给同学们提供成长自我的机会。主要工作方向包括：1、探索前沿的NLP技术：全面基于LLM的Query分析、相关性、个性化预估、满意度评估、生成式检索等，全链路应用LLM/VLM，每个细节都充满挑战；2、跨模态匹配技术：进行多模态预训练，在超大规模的多模态表征、匹配、生成等多个技术方向上持续突破，打造世界领先的多模态搜索系统；3、大规模流式机器学习技术：应用大规模机器学习，解决搜索中的推荐问题，让搜索更加个性化更加懂你；4、千亿级数据规模的架构：从大规模离线计算，分布式系统的性能、调度优化，到构建高可用、高吞吐和低延迟的在线服务的方方面面都有深入研究和创新；5、推荐技术：基于超大规模机器学习、因果推断、大模型推理等技术手段，构建业界领先的搜索推荐系统，对搜索推荐技术进行探索和创新。\n\n课题介绍：随着大模型技术的快速发展，AI搜索领域迎来了新的机遇和挑战。传统搜索技术在面对海量数据、多模态信息以及用户多轮复杂需求时，开始暴露出很多问题。因此需要基于大模型来构建下一代AI搜索系统，提升搜索系统的智能化水平，优化用户体验，具体目标包括：\n1、探索大模型与排序算法的结合，提升个性化排序的精度和用户体验；\n2、探索基于多模态预训练的端到端生成式搜索大模型；\n3、探索基于大模型Agent技术，提升复杂多义Query和多轮搜索下的用户满意度。\n\n课题挑战\n1、个性化排序的挑战：传统排序算法难以充分利用多模态信息，且模型复杂度有限，无法满足用户对精准化和个性化搜索的需求；\n2、超大规模检索排序的挑战：传统的基于判别式的级联排序系统，难以满足搜索千亿级别候选的检索排序效率需求；\n3、搜索需求日益复杂的挑战：用户搜索需求的复杂度在不断增加，传统搜索框架难以在多轮对话下，准确理解长难、多义Query的语义，导致搜索结果满意度低。\n\n课题价值\n1、技术价值：突破传统搜索技术瓶颈，构建大模型Agent驱动的下一代AI搜索架构，解决个性化排序、超大规模检索排序、复杂搜索需求理解满足等行业难题；\n2、业务价值：大幅度提升搜索的用户体验和满意度，带动搜索场景LT和主动搜索心智的提升。\n\nTopic Content\nWith the rapid advancement of foundation model technology, the field of AI-powered search is encountering new opportunities and challenges. Traditional search technologies have begun to reveal significant limitations when confronted with massive data volumes, multimodal information, and complex multi-turn user needs. It is therefore necessary to leverage foundation models to build next-generation AI search systems, enhancing the intelligence of search systems and optimizing user experience. Specific objectives include:\n1.Explore the integration of foundation models with ranking algorithms to improve personalized ranking accuracy and user experience;\n2.Explore end-to-end generative search models based on multimodal pre-training;\n3.Explore LLM-based agent technology to improve user satisfaction under complex ambiguous queries and multi-turn search scenarios.\n\nTopic Challenges\n1.Personalized ranking: Traditional ranking algorithms struggle to fully leverage multimodal information, and their limited model complexity fails to meet user demands for precise and personalized search;\n2.Ultra-large-scale retrieval and ranking: Traditional discriminative cascaded ranking systems cannot meet the efficiency requirements for retrieval and ranking across hundred-billion-scale candidate pools;\n3.Increasingly complex search needs: User search needs are growing increasingly complex. Traditional search frameworks struggle to accurately understand the semantics of long, complex, and ambiguous queries in multi-turn conversations, resulting in low search result satisfaction.\n\nTopic Value\n1.Technical value: Break through the bottlenecks of traditional search technology; build a next-generation AI search architecture driven by LLM agents; and address industry challenges including personalized ranking, ultra-large-scale retrieval and ranking, and understanding and fulfilling complex search needs;\n2.Business value: Significantly improve search user experience and satisfaction, driving improvements in search LT and users' proactive search intent.",
"requirements_text": "1、2027届毕业，获得博士学位，人工智能、计算机科学、计算机工程及相关技术专业；\n2、具备优秀的编程能力，拥有扎实的机器学习/深度学习/大语言模型专业基础，熟练掌握C/C++或Python编程语言。\n\n加分项：\n1、具备出色的问题定义、分析与解决能力；在AAAI、NeurIPS、ACL等人工智能顶会有论文发表，或具备深厚科研经验者优先考虑；\n2、有韧性，沟通能力与团队协作能力优秀；热爱技术，愿意与团队共同迎接挑战，具备创新钻研精神。\n\n1.Individuals who are completing or recently completed a PhD in AI, Computer Science, Computer Engineering, or a related technical discipline;\n2.Excellent coding skills and a solid foundation in Machine Learning/Deep Learning/LLM, proficient in C/C++ or Python.\n\nPreferred Qualifications\n1.Outstanding ability to define, analyze, and solve problems; candidates with publications at AI conferences such as AAAI, NeurIPS, ACL, etc., or in-depth research experience are preferred;\n2.Strong resilience, excellent communication and teamwork skills; passionate about technology, willing to embrace challenges with the team, and a drive for innovation.",
"skills": [AI生成
"Python",
"C++",
"机器学习",
"深度学习",
"团队协作",
"沟通能力",
"大语言模型",
"C语言"
],
"preferred_qualifications": "加分项：；在AAAI、NeurIPS、ACL等人工智能顶会有论文发表，或具备深厚科研经验者优先考虑；Preferred Qualifications；candidates with publications at AI conferences such as AAAI, NeurIPS, ACL, etc., or in-depth research experience are preferred",
"job_highlights": "未披露"
},
"contact": {
"recruiter_name": "未披露",
"contact_methods": [
{
"type": "apply_link",
"value": "https://jobs.bytedance.com/campus/position/7637133527663970613/detail"
}
]
},
"posting": {
"posting_id": "7637133527663970613",
"post_date": "2026-05-07",
"hire_time": "未披露",
"update_date": "20260917",
"job_url": "https://jobs.bytedance.com/campus/position/7637133527663970613/detail"
},
"fulltext": "基于预训练大模型的端到端生成式搜索技术研究-搜索（新加坡）Research on End-to-End Generative Search Technology Based on Pre-Trained Foundation Models-Search (Singapore)\n职位描述\n团队介绍：字节跳动搜索团队主要负责抖音、国际化短视频、今日头条、红果短剧、番茄小说、AI搜索等产品以及电商、生活服务等业务的搜索算法创新和架构研发工作。我们使用前沿的机器学习\\大模型技术进行端到端建模并不断创新突破，同时专注于分布式系统、机器学习系统的构建和性能优化，从内存、Disk等优化到索引压缩、召回、排序等算法的探索，致力于支撑字节跳动持续多元化的产品创新和高速的业务增长、基于大模型革新和重塑下一代搜索技术体系，充分给同学们提供成长自我的机会。主要工作方向包括：1、探索前沿的NLP技术：全面基于LLM的Query分析、相关性、个性化预估、满意度评估、生成式检索等，全链路应用LLM/VLM，每个细节都充满挑战；2、跨模态匹配技术：进行多模态预训练，在超大规模的多模态表征、匹配、生成等多个技术方向上持续突破，打造世界领先的多模态搜索系统；3、大规模流式机器学习技术：应用大规模机器学习，解决搜索中的推荐问题，让搜索更加个性化更加懂你；4、千亿级数据规模的架构：从大规模离线计算，分布式系统的性能、调度优化，到构建高可用、高吞吐和低延迟的在线服务的方方面面都有深入研究和创新；5、推荐技术：基于超大规模机器学习、因果推断、大模型推理等技术手段，构建业界领先的搜索推荐系统，对搜索推荐技术进行探索和创新。\n\n课题介绍：随着大模型技术的快速发展，AI搜索领域迎来了新的机遇和挑战。传统搜索技术在面对海量数据、多模态信息以及用户多轮复杂需求时，开始暴露出很多问题。因此需要基于大模型来构建下一代AI搜索系统，提升搜索系统的智能化水平，优化用户体验，具体目标包括：\n1、探索大模型与排序算法的结合，提升个性化排序的精度和用户体验；\n2、探索基于多模态预训练的端到端生成式搜索大模型；\n3、探索基于大模型Agent技术，提升复杂多义Query和多轮搜索下的用户满意度。\n\n课题挑战\n1、个性化排序的挑战：传统排序算法难以充分利用多模态信息，且模型复杂度有限，无法满足用户对精准化和个性化搜索的需求；\n2、超大规模检索排序的挑战：传统的基于判别式的级联排序系统，难以满足搜索千亿级别候选的检索排序效率需求；\n3、搜索需求日益复杂的挑战：用户搜索需求的复杂度在不断增加，传统搜索框架难以在多轮对话下，准确理解长难、多义Query的语义，导致搜索结果满意度低。\n\n课题价值\n1、技术价值：突破传统搜索技术瓶颈，构建大模型Agent驱动的下一代AI搜索架构，解决个性化排序、超大规模检索排序、复杂搜索需求理解满足等行业难题；\n2、业务价值：大幅度提升搜索的用户体验和满意度，带动搜索场景LT和主动搜索心智的提升。\n\nTopic Content\nWith the rapid advancement of foundation model technology, the field of AI-powered search is encountering new opportunities and challenges. Traditional search technologies have begun to reveal significant limitations when confronted with massive data volumes, multimodal information, and complex multi-turn user needs. It is therefore necessary to leverage foundation models to build next-generation AI search systems, enhancing the intelligence of search systems and optimizing user experience. Specific objectives include:\n1.Explore the integration of foundation models with ranking algorithms to improve personalized ranking accuracy and user experience;\n2.Explore end-to-end generative search models based on multimodal pre-training;\n3.Explore LLM-based agent technology to improve user satisfaction under complex ambiguous queries and multi-turn search scenarios.\n\nTopic Challenges\n1.Personalized ranking: Traditional ranking algorithms struggle to fully leverage multimodal information, and their limited model complexity fails to meet user demands for precise and personalized search;\n2.Ultra-large-scale retrieval and ranking: Traditional discriminative cascaded ranking systems cannot meet the efficiency requirements for retrieval and ranking across hundred-billion-scale candidate pools;\n3.Increasingly complex search needs: User search needs are growing increasingly complex. Traditional search frameworks struggle to accurately understand the semantics of long, complex, and ambiguous queries in multi-turn conversations, resulting in low search result satisfaction.\n\nTopic Value\n1.Technical value: Break through the bottlenecks of traditional search technology; build a next-generation AI search architecture driven by LLM agents; and address industry challenges including personalized ranking, ultra-large-scale retrieval and ranking, and understanding and fulfilling complex search needs;\n2.Business value: Significantly improve search user experience and satisfaction, driving improvements in search LT and users' proactive search intent.\n\n职位要求\n1、2027届毕业，获得博士学位，人工智能、计算机科学、计算机工程及相关技术专业；\n2、具备优秀的编程能力，拥有扎实的机器学习/深度学习/大语言模型专业基础，熟练掌握C/C++或Python编程语言。\n\n加分项：\n1、具备出色的问题定义、分析与解决能力；在AAAI、NeurIPS、ACL等人工智能顶会有论文发表，或具备深厚科研经验者优先考虑；\n2、有韧性，沟通能力与团队协作能力优秀；热爱技术，愿意与团队共同迎接挑战，具备创新钻研精神。\n\n1.Individuals who are completing or recently completed a PhD in AI, Computer Science, Computer Engineering, or a related technical discipline;\n2.Excellent coding skills and a solid foundation in Machine Learning/Deep Learning/LLM, proficient in C/C++ or Python.\n\nPreferred Qualifications\n1.Outstanding ability to define, analyze, and solve problems; candidates with publications at AI conferences such as AAAI, NeurIPS, ACL, etc., or in-depth research experience are preferred;\n2.Strong resilience, excellent communication and teamwork skills; passionate about technology, willing to embrace challenges with the team, and a drive for innovation.\n团队介绍：字节跳动搜索团队主要负责抖音、国际化短视频、今日头条、红果短剧、番茄小说、AI搜索等产品以及电商、生活服务等业务的搜索算法创新和架构研发工作。我们使用前沿的机器学习\\大模型技术进行端到端建模并不断创新突破，同时专注于分布式系统、机器学习系统的构建和性能优化，从内存、Disk等优化到索引压缩、召回、排序等算法的探索，致力于支撑字节跳动持续多元化的产品创新和高速的业务增长、基于大模型革新和重塑下一代搜索技术体系，充分给同学们提供成长自我的机会。主要工作方向包括：1、探索前沿的NLP技术：全面基于LLM的Query分析、相关性、个性化预估、满意度评估、生成式检索等，全链路应用LLM/VLM，每个细节都充满挑战；2、跨模态匹配技术：进行多模态预训练，在超大规模的多模态表征、匹配、生成等多个技术方向上持续突破，打造世界领先的多模态搜索系统；3、大规模流式机器学习技术：应用大规模机器学习，解决搜索中的推荐问题，让搜索更加个性化更加懂你；4、千亿级数据规模的架构：从大规模离线计算，分布式系统的性能、调度优化，到构建高可用、高吞吐和低延迟的在线服务的方方面面都有深入研究和创新；5、推荐技术：基于超大规模机器学习、因果推断、大模型推理等技术手段，构建业界领先的搜索推荐系统，对搜索推荐技术进行探索和创新。\n\n课题介绍：随着大模型技术的快速发展，AI搜索领域迎来了新的机遇和挑战。传统搜索技术在面对海量数据、多模态信息以及用户多轮复杂需求时，开始暴露出很多问题。因此需要基于大模型来构建下一代AI搜索系统，提升搜索系统的智能化水平，优化用户体验，具体目标包括：\n1、探索大模型与排序算法的结合，提升个性化排序的精度和用户体验；\n2、探索基于多模态预训练的端到端生成式搜索大模型；\n3、探索基于大模型Agent技术，提升复杂多义Query和多轮搜索下的用户满意度。\n\n课题挑战\n1、个性化排序的挑战：传统排序算法难以充分利用多模态信息，且模型复杂度有限，无法满足用户对精准化和个性化搜索的需求；\n2、超大规模检索排序的挑战：传统的基于判别式的级联排序系统，难以满足搜索千亿级别候选的检索排序效率需求；\n3、搜索需求日益复杂的挑战：用户搜索需求的复杂度在不断增加，传统搜索框架难以在多轮对话下，准确理解长难、多义Query的语义，导致搜索结果满意度低。\n\n课题价值\n1、技术价值：突破传统搜索技术瓶颈，构建大模型Agent驱动的下一代AI搜索架构，解决个性化排序、超大规模检索排序、复杂搜索需求理解满足等行业难题；\n2、业务价值：大幅度提升搜索的用户体验和满意度，带动搜索场景LT和主动搜索心智的提升。\n\nTopic Content\nWith the rapid advancement of foundation model technology, the field of AI-powered search is encountering new opportunities and challenges. Traditional search technologies have begun to reveal significant limitations when confronted with massive data volumes, multimodal information, and complex multi-turn user needs. It is therefore necessary to leverage foundation models to build next-generation AI search systems, enhancing the intelligence of search systems and optimizing user experience. Specific objectives include:\n1.Explore the integration of foundation models with ranking algorithms to improve personalized ranking accuracy and user experience;\n2.Explore end-to-end generative search models based on multimodal pre-training;\n3.Explore LLM-based agent technology to improve user satisfaction under complex ambiguous queries and multi-turn search scenarios.\n\nTopic Challenges\n1.Personalized ranking: Traditional ranking algorithms struggle to fully leverage multimodal information, and their limited model complexity fails to meet user demands for precise and personalized search;\n2.Ultra-large-scale retrieval and ranking: Traditional discriminative cascaded ranking systems cannot meet the efficiency requirements for retrieval and ranking across hundred-billion-scale candidate pools;\n3.Increasingly complex search needs: User search needs are growing increasingly complex. Traditional search frameworks struggle to accurately understand the semantics of long, complex, and ambiguous queries in multi-turn conversations, resulting in low search result satisfaction.\n\nTopic Value\n1.Technical value: Break through the bottlenecks of traditional search technology; build a next-generation AI search architecture driven by LLM agents; and address industry challenges including personalized ranking, ultra-large-scale retrieval and ranking, and understanding and fulfilling complex search needs;\n2.Business value: Significantly improve search user experience and satisfaction, driving improvements in search LT and users' proactive search intent.\n1、2027届毕业，获得博士学位，人工智能、计算机科学、计算机工程及相关技术专业；\n2、具备优秀的编程能力，拥有扎实的机器学习/深度学习/大语言模型专业基础，熟练掌握C/C++或Python编程语言。\n\n加分项：\n1、具备出色的问题定义、分析与解决能力；在AAAI、NeurIPS、ACL等人工智能顶会有论文发表，或具备深厚科研经验者优先考虑；\n2、有韧性，沟通能力与团队协作能力优秀；热爱技术，愿意与团队共同迎接挑战，具备创新钻研精神。\n\n1.Individuals who are completing or recently completed a PhD in AI, Computer Science, Computer Engineering, or a related technical discipline;\n2.Excellent coding skills and a solid foundation in Machine Learning/Deep Learning/LLM, proficient in C/C++ or Python.\n\nPreferred Qualifications\n1.Outstanding ability to define, analyze, and solve problems; candidates with publications at AI conferences such as AAAI, NeurIPS, ACL, etc., or in-depth research experience are preferred;\n2.Strong resilience, excellent communication and teamwork skills; passionate about technology, willing to embrace challenges with the team, and a drive for innovation.",
"standard_classification": {
"std_category_code": "T011",
"std_aux_codes": [
"T010"
],
"std_category_status": null,
"std_taxonomy_version": "20261104"
}
}